This study examines the targeting process of the Minimum Living Standard Scheme (MLSS, dibao) for urban residents in China. It intends to explore the factors that influence the judgment of street-level dibao workers in determining whether an applicant is eligible for the dibao; furthermore, it seeks to explain why the dibao scheme suffers from targeting ineffectiveness. During the last three decades, China has experienced both economic restructuring and social structural transformation, with many workers laid off during this transition. Poverty has gradually become a formidable challenge in urban China. The urban dibao scheme, launched in the 1990s, has become the major governmental policy response against urban poverty. After almost fifteen years of work, the dibao scheme has matured considerably; according to official statements, it now covers everyone who qualifies (yingbao jinbao), in accordance with policy. Nevertheless, academic findings suggest that China's urban dibao scheme has serious targeting ineffectiveness. This considerable issue has not been addressed sufficiently in the current literature. This study takes a bottom-up perspective and proposes that dibao workers at the street level are arguably the greatest point of leverage in policy implementation. Specifically, the "street-level bureaucracy" theory and the "implementation" theory will be employed as the basis of the explanatory framework, thereby examining how dibao workers function as street-level bureaucrats and what pitfalls result from this position. Both endogenous factors and exogenous conditions are most likely to affect street-level dibao workers' decision-making, and in turn affect the dibao scheme's targeting effectiveness. In particular, endogenous determinants, which are considered more influential than exogenous ones in this study, are street-level bureaucrats' knowledge and expertise, workload, administrative discretion, and degree of individual consensus with policy objectives. Exogenous determinants are national policy, local organizational structure and practices, and characteristics of policy clients. Based on a case study in the cities of Guangzhou and Ordos, this analytical framework contributes to our understanding of the dibao scheme's targeting ineffectiveness. The fieldwork data indicate that it is the endogenous factors of dibao workers that essentially impact on the street-level decision-making, and hence lead to targeting problems. In addition, national-level policy, local organizational practices, and the responses from dibao beneficiaries, despite affecting targeting process in different ways, have also contributed to the targeting ineffectiveness. This dissertation attempts to enrich and test the Western street-level bureaucracy theory through empirical cases from China. A caveat should be noted. The goal is to use the street-level perspective to stimulate critical thinking about the implementation deficits of social policy in China. What this study does provide is a general explanatory framework for exploring the determinants of street-level decision-making in policy implementation. Hopefully, the wider relevance of the analytical framework will serve as a vehicle for considering the likely practical response to social policies based on targeting.